This is a new audio edge memorization method I came up with. The aim of the method is to produce a sequence of audio pairs that is as short and easy to remember as possible on average, without being too hard to learn. I also wanted some properties, such as edge orientation, to be immediately obvious from the sound. Lastly, I wanted the sound as a whole to be unambiguous, that is, it should always be possible to deduce where the pairs start and end, even when pronounced together with no pause in between. I've done a lot of experimentation with different variants, and this is the final version that I'm quite satisfied with.
First of all, every edge (except the buffer) is given a consonant and a vowel. Note that I'm referring to the actual sounds, not the letters. I think most people are familiar with at least 11 consonants, so that part shouldn't be a problem (personally, I chose SFNMLRKGTDP). However, 11 vowels is a lot, and even though I'm familiar with about 20 vowel sounds because of the languages I speak, not all of them are equally good. I did a lot of experimentation, and what worked best for me was to use only the 5 Japanese vowels (AIUEO), each with two different pitches (low and high), yielding a total of 10 different vowel sounds. For the last one, I used a trick that I'll explain later.
Next, I'll explain how to make audio pairs. I like to think of it this way: For the first target, we simply pick the consonant of the corresponding edge piece if it's oriented, or the vowel if it's not oriented. Then we add the sound of the second target of opposite category, and we also add the letter H (which I'll explain in a moment) next to the vowel if the second target is not oriented. In other words, we either get a consonant + vowel (+ H) or a (H + ) vowel + consonant pair. If there's no H in the pair, it's pronounced normally. H + vowel can be pronounced normally as well, but I'll explain some optimizations later. Also, vowel + H is simply the vowel pronounced twice as long. Personally, I also change the vowel slightly into a similar but different sound. This makes sense in Swedish, but simply using the exact same vowel works fine as well.
Lastly, I'll discuss some optimizations. First of all, it's possible to remove one vowel (so that you only need 10), since whenever we have that vowel in a pair, we can simply replace it with the corresponding vowel of the consonant in the pair. This is possible because the new pair we create consists of the consonant and vowel of the same edge, which would otherwise not be used (as long as you don't memorize flips with audio pairs). Also, when pronouncing pairs together, it's possible to replace consonant + H combinations with a long version of the consonant (pronounced equally long as a short vowel), which can make the sound flow better.
Doing it this way will produce a short melody that works really well for me at least. I'm not sure if this would be possible to use in a competition though, where you have to be quiet
I think I'll stop here, but let me know if there's anything unclear about the method, or if you have anything else you'd like to say/ask about it
First of all, every edge (except the buffer) is given a consonant and a vowel. Note that I'm referring to the actual sounds, not the letters. I think most people are familiar with at least 11 consonants, so that part shouldn't be a problem (personally, I chose SFNMLRKGTDP). However, 11 vowels is a lot, and even though I'm familiar with about 20 vowel sounds because of the languages I speak, not all of them are equally good. I did a lot of experimentation, and what worked best for me was to use only the 5 Japanese vowels (AIUEO), each with two different pitches (low and high), yielding a total of 10 different vowel sounds. For the last one, I used a trick that I'll explain later.
Next, I'll explain how to make audio pairs. I like to think of it this way: For the first target, we simply pick the consonant of the corresponding edge piece if it's oriented, or the vowel if it's not oriented. Then we add the sound of the second target of opposite category, and we also add the letter H (which I'll explain in a moment) next to the vowel if the second target is not oriented. In other words, we either get a consonant + vowel (+ H) or a (H + ) vowel + consonant pair. If there's no H in the pair, it's pronounced normally. H + vowel can be pronounced normally as well, but I'll explain some optimizations later. Also, vowel + H is simply the vowel pronounced twice as long. Personally, I also change the vowel slightly into a similar but different sound. This makes sense in Swedish, but simply using the exact same vowel works fine as well.
Lastly, I'll discuss some optimizations. First of all, it's possible to remove one vowel (so that you only need 10), since whenever we have that vowel in a pair, we can simply replace it with the corresponding vowel of the consonant in the pair. This is possible because the new pair we create consists of the consonant and vowel of the same edge, which would otherwise not be used (as long as you don't memorize flips with audio pairs). Also, when pronouncing pairs together, it's possible to replace consonant + H combinations with a long version of the consonant (pronounced equally long as a short vowel), which can make the sound flow better.
Doing it this way will produce a short melody that works really well for me at least. I'm not sure if this would be possible to use in a competition though, where you have to be quiet
I think I'll stop here, but let me know if there's anything unclear about the method, or if you have anything else you'd like to say/ask about it
Last edited:
